This our work is concerned with the problem of filter design in two-dimensional (2D) discrete-time non-linear systems in Takagi-Sugeno (T-S) fuzzy model described by Fornasini-Marchesini local state-space (FM LSS) Mod...
详细信息
The exploration of high-speed movement by robots or road traffic agents is crucial for autonomous driving and navigation. Trajectory prediction at high speeds requires considering historical features and interactions ...
详细信息
ISBN:
(数字)9798350377705
ISBN:
(纸本)9798350377712
The exploration of high-speed movement by robots or road traffic agents is crucial for autonomous driving and navigation. Trajectory prediction at high speeds requires considering historical features and interactions with surrounding entities, a complexity not as pronounced in lower-speed environments. Prior methods have assessed the spatiotemporal dynamics of agents but often neglected intrinsic intent and uncertainty, thereby limiting their effectiveness. We present the Denoised Endpoint Distribution model for trajectory prediction, which distinctively models agents’ spatio-temporal features alongside their intrinsic intentions and un-certainties. By employing Diffusion and Transformer models to focus on agent endpoints rather than entire trajectories, our approach significantly reduces model complexity and enhances performance through endpoint information. Our experiments on open datasets, coupled with comparison and ablation studies, demonstrate our model’s efficacy and the importance of its components. This approach advances trajectory prediction in high-speed scenarios and lays groundwork for future developments.
Processing Earth Observation data is a compute-intensive task, due to their large size and increasing acquisition rates, often requiring the kind of computing power offered by Cloud solutions. To bridge the gap betwee...
详细信息
ISBN:
(纸本)9783030680176;9783030680169
Processing Earth Observation data is a compute-intensive task, due to their large size and increasing acquisition rates, often requiring the kind of computing power offered by Cloud solutions. To bridge the gap between the Earth data scientist and the programming expertise required to effectively use such solutions, we have developed a workflow-based process description model. It uses a description language to define algorithms in an intuitive manner, while also increasing the potential for exploiting parallelism. Within this paper, we present the results of an attempt at integrating the capabilities of the Amazon Alexa personal assistant with the process description solution that we have developed. It is meant as a first step towards a multi-modal interface that would provide support for new users to describe data processing algorithms using natural language, and potentially improve accessibility for users in the field who rely on mobile devices to visualize data.
The Internet of Things (IoT) might improve healthcare. Smart patient tracking systems may benefit. There are many networked gadgets, creating new security risks. Privacy and accessibility of patient data require stron...
详细信息
Aiming at the problems such as the lag of warehouse monitoring, the low accuracy of logistics status information acquisition and the lack of intelligence of the system, the intelligent logistics warehouse park securit...
Aiming at the problems such as the lag of warehouse monitoring, the low accuracy of logistics status information acquisition and the lack of intelligence of the system, the intelligent logistics warehouse park security perception system is designed by using artificial intelligence technology. By constructing the node model of logistics distribution control element, the link that logistics supply flows through in the operation process is analyzed, and the logistics monitoring environment is built according to the corresponding link to realize the visualization of material flow. The system can be used to simultaneously monitor the temperature, humidity, and fire conditions of the warehouse, and send the data via Zigbee technology to the computer in the management center, which processes and analyzes it. The experiment shows that the absolute error of temperature and humidity is less than 0.5℃ and less than 3% respectively. The system has reliable performance and practical value.
Today's industry is increasingly characterized by the integration of Internet of Things (IoT) devices and the rapidly spreading digitization trend, which are also known as the foundations of Industry 4.0. The impl...
详细信息
ISBN:
(数字)9798331531119
ISBN:
(纸本)9798331531126
Today's industry is increasingly characterized by the integration of Internet of Things (IoT) devices and the rapidly spreading digitization trend, which are also known as the foundations of Industry 4.0. The implementation of production automation also become an important issue, as well as the continuous collection of data and their storage in the cloud. As a result of these integration, the method of collecting data, the usability of cloud-based systems, the feasibility of artificial intelligence-based data analysis, the visualization of massive amounts of collected data, cyber security issues, etc. have become everyday issues. Industry players must devote significant attention and resources to real-time data processing to extract vital information from available datasets. This includes identifying outlier data, filtering fake information, and enabling predictive maintenance through forecasting analysis. In this article, we provide a detailed overview and the general architecture of IoT (Industrial IoT) systems, its various cloud service-based implementations, and present our alternative solution built using open-source software components. At the same time, we present some showcases and prototype applications using our system for outlier detection and predictive model building based on the collected sensor data.
The proceedings contain 251 papers. The topics discussed include: optically controlled microwave sensor for biomedical applications;deep learning network for object detection under the poor lighting condition;a brief ...
ISBN:
(纸本)9781665460842
The proceedings contain 251 papers. The topics discussed include: optically controlled microwave sensor for biomedical applications;deep learning network for object detection under the poor lighting condition;a brief analysis on machine learning classifiers for intrusion detection to enhance network security;a brief review on melanoma diagnosis models using machine learning techniques;equilibrium optimizer with deep learning model for autism spectral disorder classification;automated intracranial hemorrhage detection and classification using rider optimization with deep learning model;heart disease prediction and classification using machine learning and transfer learning model;review of machine learning algorithms for autism spectrum disorder prediction;novel approach to non-invasive blood glucose monitoring based on visible laser light;a machine learning based approach for breast cancer prediction;covid-19 infection segmentation using deep learning techniques;retinal fundus image retrieval and classification using optimal deep learning model;breast cancer segmentation by k-means and classification by machine learning;an extensive review of machine learning techniques for EEG signal processing;hybrid machine learning based false data injection attack detection and mitigation model for waste water treatment plant;twitter sentiment analysis with machine learning;hybrid particle swarm optimization with deep learning driven sarcasm detection on social media;design of kernel extreme learning machine based intelligent crop yield prediction model;and reliable densely connected network with machine learning based diabetic retinopathy grading approach.
In mobile edge computing (MEC) paradigm, for the security-critical computation tasks offloaded from the mobile device, the MEC server needs to decrypt the encrypted task data before task execution, leading to heavy co...
In mobile edge computing (MEC) paradigm, for the security-critical computation tasks offloaded from the mobile device, the MEC server needs to decrypt the encrypted task data before task execution, leading to heavy computation load on the CPUs of the MEC server. The integration of data processing unit (DPU) into the MEC server can release CPUs from data decrpytion for the purpose of reducing computation and energy overheads. This paper studies a task offloading problem in a MEC system with integrated DPUs, which aims at reducing the total energy consumption under the deadline constraint on task completion time. We formulate the studied problem as an combinatorial optimization model and propose a group mapping-based cuckoo search (GMCS) metaheuristic algorithm to explore high-quality energy-efficient solutions to the formulated model. The proposed GMCS introduces a group mapping operator and a greedy-based task offloading strategy for converting cuckoo individuals into offloading solutions and evaluating the objective values of the converted solutions, respectively. We provide a theoretical analysis to justify that the mapping operator can improve the diversity of the converted solutions and in turn the metaheuristic’s searching capability. We create various testing instances for a MEC system equipped with DPUs to demonstrate the proposed GMCS produces high-quality solutions to the studied offloading problem, with reduced energy consumption compared with baseline algorithms.
There are many advanced concepts and abstractions involved in signal-class courses, which require complex mathematical formulas to be deduced and proven. The derivation process of these formulas is difficult to follow...
详细信息
This contribution proposes a multi-modal system for older adults' health monitoring within the smart home framework that integrates Mixed Reality, a network of IoT sensors, wearable health sensors, and users' ...
详细信息
ISBN:
(数字)9798350378009
ISBN:
(纸本)9798350378016
This contribution proposes a multi-modal system for older adults' health monitoring within the smart home framework that integrates Mixed Reality, a network of IoT sensors, wearable health sensors, and users' feedback to collect data and monitor older adults' health. It provides a comprehensive framework for older adults' health monitoring and safe exercise sessions to enhance their physical and cognitive well-being. The proposed system allows older adults to control the smart home features and benefit from assistive services via a mixed reality application. The system explores exploiting machine learning models to provide personalized physical and cognitive training routines to enhance older adults' quality of life, autonomy, and general well-being. It leverages historical data coming from wearable sensors, smart mixed reality glasses, IoT sensors, and users' feedback to predict the users' health status and activity level and suggest the proper exercise routine accordingly.
暂无评论